Warspite
Full Member
Sky of blue/And sea of green
Posts: 230
|
Post by Warspite on Oct 9, 2019 8:09:40 GMT -6
I bought UA: Dreadnoughts and I'm happy with it. The graphics are very nice. It won't replace RTW 2 but it will complement it nicely. I'm glad both games exist.
|
|
|
Post by dorn on Oct 9, 2019 10:33:22 GMT -6
I agree with tortugapower ; the strategic layer of the game is very bare-bones, compared to something like Hearts of Iron. It's good at what it does, it just doesn't do much. For me, a pretty ideal game would be to combine the tactical combat and ship design of RTW with the strategic map of something like Victory at Sea. Even something like the ability to arrange your OOB before a battle would be so much better than having the game randomly assign divisions and then having to deal with that arbitrary, unrealistic handicap. I would not compare it to HoI. Strategic layer of HoI seems very nice, but AI is completely incopetent. Just play several years as UK than switch to eg. Germany and you will find that AI is lacking quite a lot. So if you play against AI it ruins the game completely as there is no challange when you get so advanced. Strategic layer in RTW is quite abstracted but make the point and AI is still quite competent even if certainly cannot be as good as player. Just think a little if you would like to do strategic layer more in depth what you need to start to simulate. You will find that you need take into considaration another and another things, make them work well together and on top of that you need to make AI competitive but as the choices increased, so the complexity and making AI competent in similar way starts to be impossible. Just look at chess and go and compare how well AI is compared to human player. More options means it is much more difficult for AI. We can wait and see what UA:D will offer in this aspect.
Even if USS Iowa meets HMS Dreadnought there still should not be 10 km difference in visibility (if radar is taken into considaration but it is not visual sight - especially from crows nest position). And especially not after USS Iowa opened fire after visual identification.
There is certainly some difference, but generally HMS Dreadnought should has higher basic chance to see USS Iowa than vice-versa (HMS Iowa has much large top than HMS Dreadnought). If you take smoke into considaration that it will be USS Iowa seeing HMS Dreadnought first, but certainly not by kilometres if crews are same quality.
|
|
|
Post by christian on Oct 9, 2019 11:38:49 GMT -6
I agree with tortugapower ; the strategic layer of the game is very bare-bones, compared to something like Hearts of Iron. It's good at what it does, it just doesn't do much. For me, a pretty ideal game would be to combine the tactical combat and ship design of RTW with the strategic map of something like Victory at Sea. Even something like the ability to arrange your OOB before a battle would be so much better than having the game randomly assign divisions and then having to deal with that arbitrary, unrealistic handicap. I would not compare it to HoI. Strategic layer of HoI seems very nice, but AI is completely incopetent. Just play several years as UK than switch to eg. Germany and you will find that AI is lacking quite a lot. So if you play against AI it ruins the game completely as there is no challange when you get so advanced. Strategic layer in RTW is quite abstracted but make the point and AI is still quite competent even if certainly cannot be as good as player. Just think a little if you would like to do strategic layer more in depth what you need to start to simulate. You will find that you need take into considaration another and another things, make them work well together and on top of that you need to make AI competitive but as the choices increased, so the complexity and making AI competent in similar way starts to be impossible. Just look at chess and go and compare how well AI is compared to human player. More options means it is much more difficult for AI. We can wait and see what UA:D will offer in this aspect.
Even if USS Iowa meets HMS Dreadnought there still should not be 10 km difference in visibility (if radar is taken into considaration but it is not visual sight - especially from crows nest position). And especially not after USS Iowa opened fire after visual identification.
There is certainly some difference, but generally HMS Dreadnought should has higher basic chance to see USS Iowa than vice-versa (HMS Iowa has much large top than HMS Dreadnought). If you take smoke into considaration that it will be USS Iowa seeing HMS Dreadnought first, but certainly not by kilometres if crews are same quality.
UAD might have overdone it i dont know but for me it doesn't feel that way besides even if i cant see yet they are shooting at me that does tell me their direction and that im probably very fucked the iowa still has much better means of spotting with better stabilization and so on in addition to producing less smoke and so on in addition to that the zoom scopes used in dreadnoughts era had horrible lighting and field of view compared to later scopes and the ones we know today point still stands there are cases when one ship can get fired on yet not have spotted the ship firing on it havent had it much in game (im mainly dicking around in the unfinished campaign which has alot more balance tech wise) so i dont know if it needs tweaking i would not have a single problem with pre dreadnoughts being outspotted by several kilometers by dreadnoughts because pre dreads had horrible optics
|
|
|
Post by Adseria on Oct 9, 2019 13:02:50 GMT -6
I agree with tortugapower ; the strategic layer of the game is very bare-bones, compared to something like Hearts of Iron. It's good at what it does, it just doesn't do much. For me, a pretty ideal game would be to combine the tactical combat and ship design of RTW with the strategic map of something like Victory at Sea. Even something like the ability to arrange your OOB before a battle would be so much better than having the game randomly assign divisions and then having to deal with that arbitrary, unrealistic handicap. I would not compare it to HoI. Strategic layer of HoI seems very nice, but AI is completely incopetent. Just play several years as UK than switch to eg. Germany and you will find that AI is lacking quite a lot. So if you play against AI it ruins the game completely as there is no challange when you get so advanced. Strategic layer in RTW is quite abstracted but make the point and AI is still quite competent even if certainly cannot be as good as player. Just think a little if you would like to do strategic layer more in depth what you need to start to simulate. You will find that you need take into considaration another and another things, make them work well together and on top of that you need to make AI competitive but as the choices increased, so the complexity and making AI competent in similar way starts to be impossible. Just look at chess and go and compare how well AI is compared to human player. More options means it is much more difficult for AI. We can wait and see what UA:D will offer in this aspect. Yes, the HOI AI is terrible, but we still haven't really worked out how to build a decent AI. Make RTW as complex as HOI and the AI would probably do even worse than HOI's AI. But I wasn't comparing the AIs, I was comparing the strategy. In RTW, there are only about a dozen places where your ships can be at any given time, and time passes on a monthly basis; in HOI, there are hundreds of locations on the sea alone, and time passes an hour at a time.
Basically, my "impossible dream" is to have HOI with RTW ship design and tactical battles.
|
|
|
Post by dorn on Oct 9, 2019 13:14:23 GMT -6
I would not compare it to HoI. Strategic layer of HoI seems very nice, but AI is completely incopetent. Just play several years as UK than switch to eg. Germany and you will find that AI is lacking quite a lot. So if you play against AI it ruins the game completely as there is no challange when you get so advanced. Strategic layer in RTW is quite abstracted but make the point and AI is still quite competent even if certainly cannot be as good as player. Just think a little if you would like to do strategic layer more in depth what you need to start to simulate. You will find that you need take into considaration another and another things, make them work well together and on top of that you need to make AI competitive but as the choices increased, so the complexity and making AI competent in similar way starts to be impossible. Just look at chess and go and compare how well AI is compared to human player. More options means it is much more difficult for AI. We can wait and see what UA:D will offer in this aspect. Yes, the HOI AI is terrible, but we still haven't really worked out how to build a decent AI. Make RTW as complex as HOI and the AI would probably do even worse than HOI's AI. But I wasn't comparing the AIs, I was comparing the strategy. In RTW, there are only about a dozen places where your ships can be at any given time, and time passes on a monthly basis; in HOI, there are hundreds of locations on the sea alone, and time passes an hour at a time.
Basically, my "impossible dream" is to have HOI with RTW ship design and tactical battles.
I completely agree that being in RTW much more choices AI will be worse. But that is developers decision and in RTW it is certainly good one. You cannot build game which AI is important part and designing it without considering how AI could adapt to the choices. This is main reason why a lot of strategy games lack reasonable AI.
|
|
|
Post by oldpop2000 on Oct 9, 2019 16:02:17 GMT -6
Yes, the HOI AI is terrible, but we still haven't really worked out how to build a decent AI. Make RTW as complex as HOI and the AI would probably do even worse than HOI's AI. But I wasn't comparing the AIs, I was comparing the strategy. In RTW, there are only about a dozen places where your ships can be at any given time, and time passes on a monthly basis; in HOI, there are hundreds of locations on the sea alone, and time passes an hour at a time.
Basically, my "impossible dream" is to have HOI with RTW ship design and tactical battles.
I completely agree that being in RTW much more choices AI will be worse. But that is developers decision and in RTW it is certainly good one. You cannot build game which AI is important part and designing it without considering how AI could adapt to the choices. This is main reason why a lot of strategy games lack reasonable AI. The problem for game developers is two fold. The platform's are limited and so is the software. I worked with AI for Naval testing stations and some other projects that I am not at liberty to talk about, and they had some issues with LISP and the limited platforms. Computer programs can only do so much, they have decision points and then have to follow a path after the decision is made. AI really doesn't learn, at least not with this limited platform or like the human mind does. I think this game is very good, the AI is one of the best I've seen. Don't expect so much from the program, just play it and enjoy it.
|
|
|
Post by aetreus on Oct 9, 2019 17:31:33 GMT -6
I would not compare it to HoI. Strategic layer of HoI seems very nice, but AI is completely incopetent. Just play several years as UK than switch to eg. Germany and you will find that AI is lacking quite a lot. So if you play against AI it ruins the game completely as there is no challange when you get so advanced. Strategic layer in RTW is quite abstracted but make the point and AI is still quite competent even if certainly cannot be as good as player. Just think a little if you would like to do strategic layer more in depth what you need to start to simulate. You will find that you need take into considaration another and another things, make them work well together and on top of that you need to make AI competitive but as the choices increased, so the complexity and making AI competent in similar way starts to be impossible. Just look at chess and go and compare how well AI is compared to human player. More options means it is much more difficult for AI. We can wait and see what UA:D will offer in this aspect. Yes, the HOI AI is terrible, but we still haven't really worked out how to build a decent AI. Make RTW as complex as HOI and the AI would probably do even worse than HOI's AI. But I wasn't comparing the AIs, I was comparing the strategy. In RTW, there are only about a dozen places where your ships can be at any given time, and time passes on a monthly basis; in HOI, there are hundreds of locations on the sea alone, and time passes an hour at a time.
Basically, my "impossible dream" is to have HOI with RTW ship design and tactical battles.
AI is hard, but generally HOI AI sucks because they don't invest enough time and money into making it competent. It's riddled with bugs, and even after years of development they still end up fixing basic things about it repeatedly. Things that should come up and be fixed in testing aren't, and I think the reality is that PDX either doesn't know how to or doesn't care to develop the AI properly.
|
|
|
Post by cogsandspigots on Oct 10, 2019 8:46:55 GMT -6
I think the biggest reason UA:D would never replace RtW for me, only supplement it, is the lack of true creative freedom in UA:D. In RtW, you can make some really strange and unique ships, but the preset hulls you have in UA:D means you’re shoehorned into certain layouts.
|
|
|
Post by director on Oct 10, 2019 11:28:02 GMT -6
oldpop2000 - 'and they had some issues with LISP' is a considerable understatement, yes? For those who don't know, LISP is a language based on recursion - the equivalent of adding up 1 to 100 by recursively nested calling the same routine and unwinding all of that once the answer is determined. 'Some issues with LISP' could mean anything from a computer vanishing down an infinite rabbit hole to an invasion of demonic biker Huns in a forest fire. I've had some brief exposure to LISP - people who program in it are not quite of this world LOL. Paradox games have bad naval AI. That's been true since at least EU2 and, with the possible exception of HoI4 and 'Man the Guns' it is still true. Ships don't carry out naval missions in a real-world way, combat is mostly absurd and the AI is utterly, absolutely hopeless at managing even trivial tasks (like NOT running your entire fleet into an unswept minefield over and over). I remember a long, excellent thread on the forum when EU3 was announced, lining out some excellent and easy ways to make navies behave like real-world navies of the period. The Paradox answer was to port over the EU2 naval code, which the project managers admitted they did not understand and did not have the resources to change. A lot of their games are fun to play and deliver more historically plausible results than competitors - but they absolutely stink at naval mechanics. Of course, their games also have no real lines of supply and not much weather... so there's that. One good 'cheat' for AI is to limit player choices to what the AI can handle. We hear about this a lot in discussions of game mechanics and the changes thereof - such and such can't be done because the AI will be exploited since it can't be 'taught' how to use those options. RtW does this with pre-set design templates the AI can put its latest equipment into, which does result in a lot of similar-looking ships (and encourages the AI to waste a lot of tonnage) but it does give the AI an easy and effective way to design ships. The restricted choices in UAD probably stem from that, and from the amount of resources required to create more parts. Modders can manage many things, but intricate and detailed 3D models are usually not among them. Back in the day, SSG made some outstanding tactical/operational level games (WW2, American Civil War) wherein the player gave orders to higher-level units like divisions and corps and the AI managed all of the brigade/regimental units (in their Russian campaign game, divisions). It was amazingly wonderful, frustrating and felt more realistic than most games I've played since... the human commander was concerned with upper-level decisions, recovering from the bone-headed errors and exploiting the occasional brilliance of the tactical commanders. Rather like playing rear-admiral or admiral mode in RtW, in fact.
|
|
|
Post by tortugapower on Oct 10, 2019 12:06:49 GMT -6
Thanks director , that was a fun post to read.
|
|
|
Post by mycophobia on Oct 10, 2019 12:30:56 GMT -6
I think the biggest reason UA:D would never replace RtW for me, only supplement it, is the lack of true creative freedom in UA:D. In RtW, you can make some really strange and unique ships, but the preset hulls you have in UA:D means you’re shoehorned into certain layouts. I think that's the disadvantage of having more things "visible" via 3D model rather than abstracted. While UA technically offer more choice in factors like bulkheads, shells, powder, etc, The fact that the main ship hulls/towers are limited to pre-set design(which in turn impact placements of guns/secondaries) does give a much stronger impression of limitation. I would've wished that the UA:D dev would've stuck to their modular block based system for hull building, but they seem to have dropped it for accessibility reasons. Then again, they aren't aiming to build a naval realism oriented FtD in the first place xD
|
|
|
Post by aetreus on Oct 10, 2019 15:44:26 GMT -6
I think the biggest reason UA:D would never replace RtW for me, only supplement it, is the lack of true creative freedom in UA:D. In RtW, you can make some really strange and unique ships, but the preset hulls you have in UA:D means you’re shoehorned into certain layouts. I think that's the disadvantage of having more things "visible" via 3D model rather than abstracted. While UA technically offer more choice in factors like bulkheads, shells, powder, etc, The fact that the main ship hulls/towers are limited to pre-set design(which in turn impact placements of guns/secondaries) does give a much stronger impression of limitation. I would've wished that the UA:D dev would've stuck to their modular block based system for hull building, but they seem to have dropped it for accessibility reasons. Then again, they aren't aiming to build a naval realism oriented FtD in the first place xD I've spent some time thinking about how a non-voxel building system for hulls and components would work in practice. Voxel is cool and all but building anything reasonably big pretty quickly becomes an absolute time sink. Volume-based with heavy automation would probably work, but the issue is getting the automation to be useful enough to support a player.
|
|
|
Post by rimbecano on Oct 10, 2019 16:11:25 GMT -6
Back in the day, SSG made some outstanding tactical/operational level games (WW2, American Civil War) wherein the player gave orders to higher-level units like divisions and corps and the AI managed all of the brigade/regimental units (in their Russian campaign game, divisions). It was amazingly wonderful, frustrating and felt more realistic than most games I've played since... the human commander was concerned with upper-level decisions, recovering from the bone-headed errors and exploiting the occasional brilliance of the tactical commanders. Rather like playing rear-admiral or admiral mode in RtW, in fact. What I'd really like to see is a game like this where the orders the player can give include defining new orders, and the algorithm AI players use to give orders is written in the same language that the human player uses to define orders. That way, part of the strategy of the game is writing friendly AI, and players that feel underchallenged can mod the enemy AI to be better.
|
|
|
Post by oldpop2000 on Oct 10, 2019 18:36:36 GMT -6
oldpop2000 - 'and they had some issues with LISP' is a considerable understatement, yes? For those who don't know, LISP is a language based on recursion - the equivalent of adding up 1 to 100 by recursively nested calling the same routine and unwinding all of that once the answer is determined. 'Some issues with LISP' could mean anything from a computer vanishing down an infinite rabbit hole to an invasion of demonic biker Huns in a forest fire. I've had some brief exposure to LISP - people who program in it are not quite of this world LOL. Paradox games have bad naval AI. That's been true since at least EU2 and, with the possible exception of HoI4 and 'Man the Guns' it is still true. Ships don't carry out naval missions in a real-world way, combat is mostly absurd and the AI is utterly, absolutely hopeless at managing even trivial tasks (like NOT running your entire fleet into an unswept minefield over and over). I remember a long, excellent thread on the forum when EU3 was announced, lining out some excellent and easy ways to make navies behave like real-world navies of the period. The Paradox answer was to port over the EU2 naval code, which the project managers admitted they did not understand and did not have the resources to change. A lot of their games are fun to play and deliver more historically plausible results than competitors - but they absolutely stink at naval mechanics. Of course, their games also have no real lines of supply and not much weather... so there's that. One good 'cheat' for AI is to limit player choices to what the AI can handle. We hear about this a lot in discussions of game mechanics and the changes thereof - such and such can't be done because the AI will be exploited since it can't be 'taught' how to use those options. RtW does this with pre-set design templates the AI can put its latest equipment into, which does result in a lot of similar-looking ships (and encourages the AI to waste a lot of tonnage) but it does give the AI an easy and effective way to design ships. The restricted choices in UAD probably stem from that, and from the amount of resources required to create more parts. Modders can manage many things, but intricate and detailed 3D models are usually not among them. Back in the day, SSG made some outstanding tactical/operational level games (WW2, American Civil War) wherein the player gave orders to higher-level units like divisions and corps and the AI managed all of the brigade/regimental units (in their Russian campaign game, divisions). It was amazingly wonderful, frustrating and felt more realistic than most games I've played since... the human commander was concerned with upper-level decisions, recovering from the bone-headed errors and exploiting the occasional brilliance of the tactical commanders. Rather like playing rear-admiral or admiral mode in RtW, in fact. I have some experience with LISP. When the Navy was developing the testing system for the F-18A, they had an automatic tester that was going to use LISP to enhance troubleshooting. So, NAVAIR sent the programmers to me, and they spent a week watching every move I made on the VAST station. Every time I looked at a book, program listing, they asked the question. They said that my troubleshooting methods were going to be the basis for the test sets. It drove me to drink, having to think and explain every move I made. Anyway, I don't know how it turned out but I could have strangled my engineering buddies who did that to me. They did give me some data on how LISP worked. I remember SSG games, I think I had one. Carriers at War, I believe.
|
|
|
Post by gabeeg on Oct 10, 2019 23:51:15 GMT -6
I am looking forward to UA:D it looks intriguing and visually appealing (Thanks for the Youtube vids Tortuga!). I have a feeling RTW2 will remain my primary go-to but we shall see. I am in the camp that RTW2 does not need 3D to improve it...but I would like to see the 2D improved and modding of 2D ships easier and more cohesive. I always have room for a non-arcade naval war game (WoW looks great...but holds no interest for me), if they make the I will come.
|
|